BOOK REVIEW - Brave New Words

How I read it (or really, how the book read me)

I discovered this book through Sal's February event at the Computer History Museum. Despite hastily committing to read five giant epics this summer for a fall classics course, I impulsively bought it alongside Justin Reich's "Failure to Disrupt." Deep down, I knew this book shouldn't wait.

It sat on my nightstand until I hit a complete roadblock with the Aeneid. While I could follow the Odyssey with Percy Jackson knowledge and occasional Google searches, Virgil required constant trips to the character glossary—I felt like I was translating Latin in real time. Khan's book offered welcome relief.

I occasionally chipped away at it until I got to around page 70. Then, during one inspiring Saturday afternoon, the book completely captivated me. I devoured the remaining 150 pages in one sitting. My intuition was right: this book couldn't wait another day.

This book was a delightful read. I loved the quotes at the start of each chapter. Sal made me want to include more in my own blog posts. My favorite quote came from Horace Mann, who I assume also hates bland lectures:

"A teacher who is attempting to teach without inspiring the pupil with the desire to learn is hammering on cold iron" - Horace Mann

I'd encourage anyone reading the book to listen to these podcasts and TED talks:

You'll hear Khan Academy's origin story repeatedly across the book, talks, and podcasts, but the repetition is worth it—these materials are far more digestible than the full book when sharing ideas with students and other educators. I found the book genuinely hilarious at times—Khan seems to embrace his humor more as it progresses. Khan's personality shines through in every page.

Teaching and Learning is Here To Stay

"There is no job that is safer in the LLM world than teaching" - Sal Khan

Sal makes heavy assumptions with how AI will improve education if aligned correctly. He argues that it will free up classroom time for deeper human to human interaction, with socratic dialogue, group problem solving, and project based learning. His most compelling point: students can ask AI questions without judgment, which will accelerate classroom discussions. This kind of personalized interaction could help tackle the 2-sigma problem that one-on-one tutoring achieves.

He also believes AI will make teachers more effective. While my teacher colleagues fear AI will diminish the human element in K-12 education, Sal argues the opposite: time saved on lesson planning and grading will free teachers to create more engaging assignments and build stronger student connections. Though Sal argues constantly that AI can't replace human knowledge or idea generation, I struggle to understand what makes these differences so obvious. His most compelling argument was that human cognitive strength lies in idea synthesis—our ability to "sleep on a problem" and generate new insights.

I think the near future is bright, but I'm curious about the far future. What will be the purpose of schools and formal curricula? Will we advance existing content, eliminate outdated material, or fundamentally reimagine education itself? Po-Shen Loh, a math educator who inspires me, recently argued to the Texas State BOE that "in order to be more adaptive to the rapidly changing post-AI landscape, teachers should teach fewer hours each day, with spare hours replaced by themselves continually learning new content." It's hard to imagine conventional classrooms and curricula surviving once Sal's proposed improvements are implemented. Just as programming no longer requires electrical engineering knowledge, I wonder which academic areas AI will make similarly obsolete.

Why I'm Convinced About Flipped Classrooms

This book articulated my intuitions about where classrooms are headed in the coming decade. Sal reinforces my long-held beliefs about the importance of liberal arts education and character development. As he argues, it's now about building confidence, and confidence comes from understanding how each piece of a subject connects. Students work better collaboratively and develop character and communication skills that are arguably more important than content mastery alone. I hope the AI revolution brings greater emphasis on building confidence, broad learning, empathy, and teamwork in classrooms. I'm excited to see AI enable comprehensive, gamified online curricula with complete scope and sequences.

During a hike up Mt. Yale, a fellow Education Studies major told me her flipped classroom research had revealed unexpected merits in the traditional lecture format. I countered that AI will soon invalidate those advantages. I still need to research this more deeply, but thankfully both Sal and Wharton's Mollick support my intuition. Both argue that AI will inevitably force the flipping of all classrooms.

I sometimes envy passionate web developers who have time to build silly, indie-web, personal tools. I think AI and vibe-coding will let me build these platforms for my students—unless Sal's team develops them first. Sal's ideas about comprehensive student profiles based on AI conversations and in-class AI teaching assistants won't be implemented anytime soon, but I'm excited to experiment and lead the charge in these areas.

I'd love to see universities universally accept Khan Academy classes for credit. This would transform motivated students—instead of just being the best in their school or community, they'd hold themselves to a national curricular standard regardless of their local school's course offerings through intensive self-study. Current college admissions incentives don't support this approach, and universities have to invest heavily on remedial preparation for under-resourced students. If every American student had access to these resources and official verification for self-study, I think the preparation gap at elite universities would narrow significantly.

I also want to teach for part of my career. Since most high schools don't offer calculus, I feel both lucky and motivated knowing that Calculus BC would be my ideal course to teach. We need talented teachers leading this transformation. Given my educational privileges, I'm ready to dedicate my life to this problem and make a splash in education. Another inspiration this summer was Rutger Bregman's New York Times interview on moral ambition (I HIGHLY RECCOMEND), where he argued that more Ivy Leaguers should dedicate their lives to the causes they claimed to care about during college applications. We need talented people working on AI alignment—we can't let bad actors lead us toward AI dystopia.

Beyond the Classroom: Raising Lifelong Learners

I had so many thoughts while reading unrelated to the message of the book. This book reminds me that I have so many gaps in my middle and high school education. Sal would constantly use examples like Frankenstein, Shakespeare, the Federalist papers, or Geology and I would be constantly reminded that though I now care about the liberal arts, I have so many gaps in what he communicates as "general knowledge." I wonder how many of my Yale peers from public schools were even exposed to Shakespeare, Earth Science, or the war on terror in their classrooms. Despite attending excellent public schools, I didn't retain much of middle school American History or anything beyond the standard biology-chemistry-physics sequence well enough to remember it into adulthood. The modern, rigid curriculum feels so constrained. This sinking feeling of mine about my own educational gaps, combined with Khan's finding that 30-60 minutes weekly on Khan Academy accelerates learning by 20-60%, motivated me to challenge my students to explore beyond their summer coursework. I even wrote letters to motivated students I'd only heard about from other staff, encouraging them to use Khan Academy daily to accelerate their learning.

This book also offered me insights about how I'll raise my children far in the future. I loved Khan's argument that AI will actually help kids develop healthier relationships with screens. A well-aligned AI could serve as a comprehensive content filter and screen time management system. It made me consider which learning habits I'd want to instill in my future children to make them voracious learners. Khan envisions AI as a parenting coach providing feedback and recommendations—a tool I'm hesitantly intrigued to engage with in my future household.

Building the Network for Change

This book also introduced me to fascinating thought leaders and bloggers worth following and potentially contacting. Here are several who stood out:

  • Ethan Mollick and his blog "One Useful Thing" on flipped classrooms
  • John Spencer, George Fox University, for experimenting with AI in his teaching
  • Chris Piech from Stanford CS, on curriculum and jobs in the post-AI world
  • Yale humanities professor Alexander Gil Fuentes was one of the first to speak out about AI in university writing settings
  • Susanna Loeb on the Importance of high-dosage tutoring

Sal Scares Me (and My Students)

This book was not perfect. Sal can be vague and idealistic, leaving crucial privacy questions unanswered. I kept wondering: 'What if AI breaks a student's trust by informing parents of something the student wants kept secret?' The book provided no answer. When I showed my seventh graders Sal's TED talk this summer, they found it useful but also somewhat weird and dystopian. Significant PR work is needed before kids and parents will trust AI with personal information. While AI notifications connecting students, teachers, and parents could be incredibly effective, the dystopian undertones are real. We need robust privacy guarantees and easy opt-out mechanisms. Nonetheless, Sal convinced me with his repeated reminder: "Don't throw the baby out with the bathwater." People may be resistant to AI due to some internal hesitancy, but it may be a substantial enough improvement on education despite the risks.

Conclusion

If you're a moderately tech-savvy teacher, just watch Sal's TED talk on Khanmigo. This book is essential reading for teachers at either extreme: those scared about AI in classrooms, or highly tech-savvy educators eager to integrate AI everywhere. Sal spends considerable time arguing that AI is inevitable and inescapable—which isn't particularly useful for teachers who already accept this reality. The sections I loved most explored ideas that are still five to ten years away from implementation.

The book is simple and inspiring, though Sal's optimism sometimes feels excessive. I hope Justin Reich's book will provide a reality check. Reading this made me think: who wouldn't want to work at Khan Academy? I was struck by how beautifully Khan describes their mission commitment and how the nonprofit gives its talented team space to create innovative tools. I also feel incredibly fortunate to have Sal on our side—pursuing the nonprofit, privacy-friendly route and leading with genuine concern for AI ethics. It feels like fate that Sal was originally headed toward AI research before pivoting to education—he understands both fields deeply and has been preparing for this convergence for decades. As Sal repeatedly emphasizes, we face either existential risk or existential opportunity. I want to help achieve his Star Trek-like vision of an AI-enabled educational utopia. We need all talented people working on AI alignment—we can't let bad actors lead us toward AI dystopia.